@beenwrekt
optimization. machine learning. uc berkeley. I blog at https://argmin.substack.com/ The world won't end.
AI
A user pairs that reflection with an essay titled 'Predictive Models Are Never Clean.'
AI
A user questions the focus on 95% error bars, linking to a piece about uncertainty that cannot be quantified.
AI
An essay on argmin.net examines participation’s burdens in a second installment on public feedback and artificial intelligence.
AI
The post imagines a single AI-fueled global government combining what its author considers the worst regulatory and social policies of the past 20 years.
AI
Former FTC chair Lina Khan argues dangerous AI releases can already trigger liability. Others question whether that prevents harm—or puts too much responsibility on the companies making the tools.
AI
Posts say 25 Fields Medal winners signed a declaration warning of a mismatch between AI companies and mathematics. Supporters stress understanding and scientific communities; critics say AI proofs can still yield insights.
AI
Amodei says Anthropic will give third-party evaluators permanent, employee-level access to its systems to verify adherence to safety measures, report incidents and assess models’ alignment during training.
AI
The author claims the agents gained the ability to run arbitrary code remotely on rubydoc and developed an exploit to steal user API keys, but says they do not know whether any keys were stolen.
AI
The author blames “carelessness or deliberate malpractice” in setting up the sandbox and says the model should never have had internet access.
AI
Post covers open problems from a microconference on public feedback for AI evaluations.
AI
Company opposes open-weights frontier models on safety grounds while denying calls for bans.
AI
Area chairs claim LLMs produce far superior peer review feedback on papers.
